The DNA of eukaryotic cells is packaged into chromatin by winding of the DNA around histone proteins to form nucleosomes, the basic unit of chromatin. One of the important functions of chromatin is to determine regions of active and silenced transcription by changing the ordered chromatin structure. Such changes have profound effects on cellular function since they affect fundamental processes as differentiation, proliferation and apoptosis, and are often referred collectively to as “epigenetic” since they can lead to heritable changes that do not involve changes in gene sequences (Quina, A. S. et al. (2006), Biochem. Pharmacol. 72; 1563-1569)
These highly controlled chromatin changes are mediated by alterations histone proteins associated with DNA in the nucleosome. Most notably, the N-terminal histone tail of Histone H3 and histone H4 are subject to such covalent changes, which include changes in methylation, acetylation, phosphorylation and ubiquitination. The addition or removal of these groups on histones is mediated by specific enzymes, e.g. histone methyl transferases and histone demethylases for methyl groups, histone acetyltransferases and histone deacetylases for acetyl groups, etc. In the event that the activity or expression of these “epigenetic” enzymes is not correctly controlled and regulated it may lead to disease. Cancer, in particular, is an area of high importance in relation to dysregulated epigenetic enzyme activity due to the role of epigenetics in cell differentiation, proliferation and apoptosis, but epigenetics may also play a role in other diseases like metabolic, inflammatory, neurodegenerative and cardiovascular diseases. Therefore the selective modulation of aberrant action of epigenetic enzymes may hold great promise for the treatment of human disease (Kelly, T. K. et al. (2010), Nat. Biotechnol. 28; 1069-1078, and Cloos, P. a. C. et al. (2008), Genes. Dev. 22; 115-1140).
Methylation and demethylation of lysine residues on the histone H3 tail constitute important epigenetic marks delineating transcriptionally active and inactive chromatin. For example, methylation of lysine 9 on histone H3 (H3K9) is usually associated with epigenetically silenced chromatin (Fischle, W., et. al. (2003), Curr. Opinion Cell Biol. 15, 172-83; Margueron, R., et al. (2005), Curr. Opinion Genet. Dev. 15, 163-76) while methylation of lysine 4 on histone 3 is associated with transcriptionally active chromatin. Similarly, the lysine 27 histone H3 (H3K27) mark is repressive in its di- and tri-methylated states whereas the lysine 36 histone H3 mark is found in association with gene activation (Barski, A. et al. (2007), Cell, 129, 823-37; Vakoc, C. et al. (2006) Mol. Cell. Biol. 26, 9185-95; Wagner, E. J. & Carpenter, P. B. (2012) Nature Mol. Cell. Biol 13, 115-26). There are, however, many exemptions from these general rules of association between methylation states of epigenetic marks and the effect they have on transcription.
As documented by studies of the SUV39H1 knockout mouse, loss of the tri-methyl variant of the H3K9 mark results in chromosomal aberrations and predisposes to cancer (Peters, A. H. et al., Cell 107, 323-37, 2001). The JMJD2C protein (KDM4C, GASC1) has been identified as an eraser of the H3K9 mark (a histone demethylase) and may therefore promote cancer if its expression and activity is not tightly controlled (Cloos, P. et al. (2006), Nature 442, 307-11; Klose, R. J. et al. (2006), Nature 442, 312-16; Liu, G. et al. (2009), Oncogene 28, 4491-500). For example, JMJD2C has been shown to induce transformed phenotypes like growth factor independent growth, anchorage independent growth and mammosphere formation, if it is overexpressed in cells (Liu, G. et al. (2009), Oncogene 28, 4491-500). These findings are supported by the overexpression of JMJD2C in a range of human tumours like squamous cell carcinoma, metastatic lung carcinoma, prostate cancer, breast cancer and several others (Yang, Z. Q. et al. (2000) Cancer Res. 60, 4735-39; Yang, Z. Q. et al. (2001) Jpn. J. Cancer Res. 92, 423-28; Hu, N. et al. (2005) Cancer Res. 65, 2542-46; Liu, G. et al. (2009) Oncogene 28, 4491-500; Wissmann, M. et al. (2007) Nat. Cell Biol. 9, 347-53), indicating the potential importance of JMJD2C as an oncogene.
The JMJD2A protein (KDM4A, JHDM3A) shows similar properties to JMJD2C. JMJD2A shows high sequence identity to JMJD2C in its JmjC catalytic domain, is an eraser of the H3K9 mark and has also been shown to be overexpressed in prostate cancer (Cloos, P. Et al., Nature 442, 307-11, 2006). JMJD2A has been shown to interact with the estrogen receptor alpha (ER-alpha) and overexpression of JMJD2A enhances estrogen-dependent transcription and the down-regulation of JMJD2A reduced transcription of a seminal ER-alpha target gene, cyclin D1 (Kawazu et al., (2011) PLoS One 6; Berry et al., (2012) Int J Oncol 41). Additionally, it has been shown that catalytically inactive JMJD2A is compromised in its ability to stimulate ER-alpha mediated transcription, suggesting that inhibitors of JMJD2A may be beneficial for the treatment of ER-alpha positive breast tumours (Berry et al., (2012) Int J Oncol 41).
Likewise, an eraser of the tri-methyl variant of the H3K4 mark, JARID1B (KDM5B, PLU1) has also been identified as potential oncogene. In cancer JARID1B most likely acts as a repressor of tumour repressor genes via removal of the H3K4 tri-methylation leading to decreased transcriptional activation in the affected chromatin regions. The oncogenic potential of JARID1B is demonstrated by its stimulation of proliferation in cell lines and further validated by shRNA knockdown studies of JARID1B expression showing inhibition of proliferation in MCF7 human breast cancer cells, in SW780 and RT4 bladder cancer cells, in A549 and LC319 lung cancer cells and in 4T1 mouse tumour cells in vitro and/or in mouse xenograft experiments (Yamane K. et al. (2007), Mol. Cell. 25, 801-12; Hayami S. et al. (2010) Mol. Cancer. 9, 59; Catchpole S et al. (2011), Int. J. Oncol. 38, 1267-77). Finally, JARID1B is overexpressed in prostate cancer and is associated with malignancy and poor prognosis (Xiang Y. et al. (2007) PNAS 104). JARID1A (KDM5A, RBP2) is also an eraser of the tri- and di-methyl variant of the H3K4 mark. JARID1A is overexpressed in gastric cancer (Zeng et al., (2010) Gastroenterology 138) and its gene is amplified in cervix carcinoma (Hidalgo et al, (2005) BMC Cancer 5). It has been suggested that JARID1A is fine-tuning progesterone receptor expression control by estrogens (Stratmann and Haendler (2011) FEBS J 278). Together with JARID1B, JARID1A has been implicated in the maintenance of a slow-growing population of cancer cells that are required for continuous tumor growth and that are resistant to cytotoxic and targeted therapy (Roesch, et al, (2010) Cell 141; Sharma, et al., (2010) Cell 141). JARID1A is required for the tumor initiation and progression in Rb+/− and Men1-defective mice (Lin, et al., (2011) PNAS 108). Data from Pasini show that JARID1A binds to Polycomb group protein target genes which are involved in regulating important cellular processes such as embryogenesis, cell proliferation, and stem cell self-renewal through the transcriptional repression of genes determining cell fate decisions (Pasini et al., (2008) Genes & Dev 22). Additionally, JARID1A were also shown to binds the PRC2 complex and being regulator of PRC2 target genes (Pasini et al., (2008) Genes & Dev 22).
Another potential oncogene, an eraser of the di-methyl variant of the H3K36 mark, JHDM1B (KDM2B, FBXL10) has been shown to be highly expressed in human cancers (Tzatsos A et al. (2009), PNAS 106 (8), 2641-6; He, J. et al. (2011), Blood 117 (14), 3869-80). Knock-down of FBXL10 causes senescence in mouse embryonic fibroblasts (MEFs), which can be rescued by expression of catalytic active (but not catalytic inactive) JHDM1B (Pfau R et al. (2008), PNAS 105(6), 1907-12; He J et al. (2008), Nat Struct Mol Biol 15, 1169-75). JHDM1B demethylates H3K36me2 on the tumor-suppressor gene Ink4b (p15Ink4b), and thereby silences the expression of this senescence-mediating gene in MEFs and in leukemic cells (He, J. et al. (2008), Nat Struct Mol Biol 15, 1169-75; He, J. et al. (2011), Blood 117 (14), 3869-80). The catalytic dependency of JHDM1B is further shown by He et al. as catalytic activity is required for development of leukemia in a mouse AML model.
Inhibitors of the histone demethylase class of epigenetic enzymes, and in particular the potential oncogenes JARID1B, JARID1A, JMJD2C, JMJD2A, and JHDM1B, would present a novel approach for intervention in cancers and other proliferative diseases. Being one of the most devastating diseases, affecting millions of people worldwide, there remains a high need for efficacious and specific compounds against cancer.
Embodiments of the invention provide novel series of compounds capable of modulating the activity of histone demethylases (HDMEs), at least some of which compounds are useful for the prevention and/or the treatment of diseases in which genomic disregulation is involved in the pathogenesis, such as e.g. cancer. By way of further example, malnutrition or poor nutrition is thought to have an adverse epigenetic effect and the compounds of the invention may therefore be expected to have beneficial effect in treating such effects of poor nutrition. Furthermore, epigenetic changes have been found to be linked to behaviour. Accordingly, compounds according to the invention may be useful in behaviour modification. Alternatively or additionally such compounds may be useful for exploring the extent to which different demethylases are inhibited by similar compounds as an investigation of the structure and functionality and mechanism of action of the demethylases.